(This is my first long-form in nostr, don’t bully me)

I’ve been exploring the rabbit hole of the zettelkasten for some time now. I was somewhat amazed about the creator of the zettelksten’s (Luhmann) ability to retain so much information systematically and to write so many books and articles, so I gave it a try. I also found the conversations of @npub1...jc2jl and @npub1...qa5sf about biology and mycelial networks very convincing and relatable.

I have never returned from that journey…

The beginning

Graphical complexity:

I was in a very weird spot. Some time ago a guy was talking about a “more correct” way of doing a zettelkasten and showed an article about that, but I didn’t read everything in that article because it was pretty long and I am not pretty good at reading in silence (a defect I’ve been trying to overcome by reading English aloud). Then, I was busy and forgot where the article was. Something I remembered, however, were the rules of making every Zettel with a limited number of words, putting a unique number that identified each of them and thei position inside the tree; and making a completely new Zettel when you wanted to modify an already existing one. The big bodies of text could be decomposed into smaller specific Zettels

With those tools, I started making 2 zettelkastens, one about clean code (because there were too many ideas in that subject and I didn’t understand them appropriately) and other about food. When trying to understand how to categorize the different recipes and the different types of individual ingredients, I suddenly hit a roadblock trying to decide how to connect them so they could be numbered appropriately. “This category goes above, these ingredients go below, these emergent connections go here and there…” It was a total mess and the graph didn’t make any sense to me. At the same time, given that I was primed with the idea of complex systems that evolve from simple systems, I was trying to find a way of reducing the complexity of the different elements of the graph because whatever.

In the middle of that juggling was when my first moment of Zett serendipity hit. I don’t need any hierarchical tree structure of categories and topics. The emergent connections will be the only way of making my zettelkasten work, not higher nor lower ideas. Every part will be equally conquered by emergent order and those connections will at the same time diminish the complexity of the graph.

This is the definition of graphical complexity that I drew at the time:

“A zettelkasten’s graphical complexity is the complexity that results from segmenting and choosing a specific part of it. It’s expressed in how the different components of the extracted segment behave together It’s inversely correlated to the amount of “emergent” links between zettels”

With that definition, I knew something big was coming…

(Another thing I wanted to say in this part is that I drew inspiration from nature when I tried to tackle the problem of graphical complexity. I thought that the typical spider web was sweet. A few concentric circles and a number of rays coming from the center of the circles towards the periphery. It doesn’t matter how many straight lines come from the center, all vertices in the web have only 4 connections except for the central vertex. If you employ only 3 concentric circles and a really high number (millions, billions, trillions) of rays, you could still walk from one vertex to another in just 6 steps. I thought that if I (magically) eliminated the central vertex, I would get the Ideally simple graph. Obviously this isn’t possible but it will prove a useful idea in the future)